Effects of Starvation and Water Quality on the ... - ACS Publications

Sep 11, 2008 - The aim of this study was to determine the effects of starvation and water quality during the purging process on the biometric paramete...
0 downloads 0 Views 133KB Size
J. Agric. Food Chem. 2008, 56, 9037–9045

9037

Effects of Starvation and Water Quality on the Purging Process of Farmed Murray Cod (Maccullochella peelii peelii) GIORGIO PALMERI,† GIOVANNI M. TURCHINI,*,† RUSSELL KEAST,† PHILIP J. MARRIOTT,‡ PAUL MORRISON,‡ AND SENA S. DE SILVA§ School of Life and Environmental Sciences, Deakin University, P.O. Box 423, Warrnambool, Victoria 3280, Australia, Australian Centre for Research on Separation Science, School of Applied Sciences, RMIT University, G.P.O. Box 2476V Melbourne, Victoria, Australia, and Network of Aquaculture Centres in Asia-Pacific (NACA) P.O. Box 1040, Kasetsart Post office, Bangkok 10903, Thailand

The aim of this study was to determine the effects of starvation and water quality during the purging process on the biometric parameters, fatty acids, and flavor volatiles of Murray cod farmed in a recirculation system. Market size Murray cod, at the end of the grow-out stage, were divided into eight treatments. The treatments were either fed/starved (F or S) and kept in clean water (CW: CWF2, CWS2, CWF4, and CWS4) or fed/starved and kept in recycled water (RW: RWF2, RWS2, RWF4, and RWS4) for either 2 or 4 weeks. Fish were sampled at 0, 2, and 4 week intervals. Food deprivation was responsible for a significant (P < 0.05) weight loss compared to that of fed treatments. The same was observed for the condition factor (K), hepatosomatic index (HSI), and dress-out percentage (DP). No significant changes were, however, observed in the visceral fat index (VFI). Saturated fatty acids (SFA) were highest in RWF4 and lowest in CWS4 (P < 0.05), while monounsaturated fatty acids (MUFA) were lowest in CWF4 (P < 0.05). Starvation did not affect the flavor volatile compounds, which were mainly affected by changes in water quality. Specifically, total aldehyde (% w/w) content was significantly (P < 0.05) affected by water quality, but the time of purging was not responsible for any noteworthy differences. This study was able to separate the effects of starvation and water quality, in the purging process, on the final eating quality of farmed market size Murray cod. It is concluded that because of the inevitable weight loss during starvation, Murray cod should be fed during the purging stage but kept in clean water and deprived of food only for the time necessary to empty the gastro-intestinal tract. KEYWORDS: Aquaculture; fatty acids; fish; off-flavor; SPME; GCMS; volatile compounds

INTRODUCTION

Purging is a common practice in aquaculture to ensure that the product reaches the market with an empty gastro-intestinal tract and without off-flavors (1). The duration of purging is dependent on many factors, mainly the species cultured, the nature of the off-flavor contaminants, intensity of the off-flavor, culture method, and environmental conditions (2). Usually, fish intensively reared at high stocking density require more time to be purged compared to fish kept at low stocking densities in semi-intensive or extensive environments. This is mainly due to the fact that under intensive rearing conditions degraded water quality due to over accumulation of * Corresponding author. Phone: +61 3 5563 3312. Fax: +61 3 55633462. E-mail: [email protected]. † Deakin University. ‡ RMIT University. § Network of Aquaculture Centres in Asia- Pacific (NACA).

feed and nutrients from fecal matter, can play a major role in contributing unpleasant taints to the flesh (3). However, it is also possible that in extensive pond conditions, environmental contaminants, namely, geosmin and 2-methylisoborneol (MIB), can seriously affect the marketability of the product (4). The purging process commonly involves moving market sized fish to clean water and starving them from a few days to many weeks (5) before they are processed and packaged, or transported live. In more recent years, a longer term purging up to 60 days and associated longer starvation has also been adopted to stabilize and improve the flesh quality (6, 7). In fact, starvation can also be considered a conditioning technique as it enhances the biochemical and microbial storage stability of the carcass. By reducing the amount of feces in the intestine, spoilage is delayed, and digestive enzyme activity is reduced. If further processing steps are considered, e.g., filleting and freezing, an interruption in the feeding before slaughter may be a determinant factor of product shelf life (8). Furthermore, purging can also

10.1021/jf801286n CCC: $40.75  2008 American Chemical Society Published on Web 09/11/2008

9038

J. Agric. Food Chem., Vol. 56, No. 19, 2008

improve the overall final nutritional qualities of farmed fish by reducing excessive fat deposition and increase in the percent content of the health-promoting long chain omega-3 polyunsaturated fatty acids (7, 9). A drawback of purging procedures is that, due to feed deprivation, there is an inevitable weight loss (1, 9), and under some circumstances, potential increased aggressive behavior can lead to external damage (such as fin and skin abrasions, which can be responsible for depreciation at the market) or even cannibalism and death. Murray cod, Maccullochella peelii peelii (Mitchell), is the largest Australian native, freshwater, carnivorous, warm water fish. Currently, Murray cod supports a small but growing aquaculture industry within Australia (10), with a production of 87.2 tons valued at just over $1.4 million in 2005/2006 (11). Most of this production is from indoor, controlled-environment, recirculating aquaculture systems (RAS) and is sold domestically. Palmeri et al. (9) found that Murray cod, on average, lose between 4% and 9% body weight after 2 and 4 weeks of purging, respectively. It was also found that, following a consumer test, Murray cod purged for 2 weeks were preferred to unpurged fish and not dissimilar to fish purged for 4 weeks. During this study, protein and hepatic reserves, not fat, either body or perivisceral, were affected by the purging procedures. As described previously, commonly implemented and studied purging procedures involve a combination of starvation and transfer of fish into clean water. No studies, to date, have investigated the individual and combined effects of feed restriction and water quality on the purging process. It is possible that by limiting starvation time to 2-3 days while maintaining good water quality, acceptable edible qualities are maintained without major weight losses. This study was therefore designed to determine if Murray cod can be purged in clean water without undergoing a prolonged regime of feed restriction past the time necessary to empty the gut. MATERIALS AND METHODS Purging Facility. Two recirculating facilities were used for the purpose of this trial. The first one was the commercial production facility in which fish have been previously grown, designed to produce 20 tons of Murray cod per year. The system is fitted with microbead biofiltration, UV sterilization, and a 40 µm drum filter for collection and waste disposal. All fish were housed in a 2500 L tank during the growout phase and then transferred, within the same system, into 6 circular 600 L tanks for the experimental purging period. The second system was a RAS purging system designed to hold 1 ton of fish at any one time. The purging set up consists of 6 circular 600 L tanks, part of a 15 tank recirculating system with a total volume of approximately 15, 000 L fitted with trickling biofiltration and sand filter for waste collection and removal. The water (dechlorinated town water) was exchanged at a rate of >2,000 L day-1. Salinity was kept between 2 and 4 g L-1 and temperature at ∼17 °C. All the other water quality parameters for both systems were adequate to the culture of this species (12) and are reported in Table 1. Experimental Fish and Sampling Procedures. Average market size fish (534.1 ( 16.3 g) were sourced from the stock of the intensive RAS facility located at Deakin University, Warrnambool, Australia. During the final stages of the grow-out period, the fish were fed a commercially extruded diet (Classic, Skretting, Tasmania, Australia; moisture ) 86.7 ( 0.3 mg g-1; crude protein ) 419.7 ( 0.1 mg g-1; total lipid ) 187.3 ( 0.0 mg g-1; ash ) 76.7 ( 0.3 mg g-1; energy ) 20.5 kJ/g). Individual weight and length of harvested fish were determined to the nearest g and cm, respectively, and 40 fish per tank were allocated to 12 purging tanks (three tanks per treatment). At the beginning of the trial fish were divided into two groups and randomly allocated to 12 tanks, 6 in the commercial facility and 6 in

Palmeri et al. Table 1. Water Quality Parameters for Systems Using Recycled and Clean Water

a

water quality parameter

recycled water

clean water

temperature °C pH DOa (mg L-1) NH3-N (mg L-1) NO2-N (mg L-1) NO3-N (mg L-1) alkalinity (mg L-1) BODa (mg L-1) salinity (gL-1) total P (mg L-1) turbidity (NTU) TSSa (mg L-1) TDSa (mg L-1)

22.6 7.8 7.7 0.5 2.5 50 430 10 0 3 18 17.2 1400

17.1 6.9 8.8 0.5 1 30 86 0 3 0.05 0 0 2750

See Abbreviations Used.

the purging facility and treatments designed as follows. Purging facility: CWF2/CWF4 ) clean water and fed for 2 weeks/4 weeks (N ) 3); CWS2/CWS4 ) clean water and starved for 2 weeks/4 weeks (N ) 3). Commercial facility: RWF2/RWF4 ) recycled water and fed for 2 weeks/4 weeks (N ) 3); RWS2/RWS4 ) recycled water and starved for 2/4 weeks (N ) 3). Treatments that were not undergoing starvation were fed (2% b/w day-1) a Skretting Classic diet by means of 12 h belt feeders. Fish were sampled at the beginning of the trial (time zero) and at weeks 2 and 4. Fish sampled at time zero were not starved, and they were regularly fed the day before culling. Three fish per replicate tank were harvested, transferred to a drum containing ice slurry, and subsequently culled by cutting the main arterial vessel in the throat and left in the ice slurry until no movement was observed. They were then removed from the ice, dried with a paper towel, gutted, gilled, filleted, and frozen at -20 °C until needed for analysis. All procedures used were approved by the Deakin University Animal Welfare Committee. Biometric and Growth Parameters. The main biometric parameters including total weight (TW), total length (TL), somatic weight (SW), liver weight (LW), fillet weight (FW), and perivisceral fat weight (PFW) were recorded (all weights were in g and length in cm). The following parameters were also calculated: Fulton’s condition factor, K ) (TW ÷ L3) × 100; hepatosomatic index, HSI (%) ) (LW ÷ TW) × 100; visceral fat index, VFI (%) ) (PFW ÷ TW) × 100; dress-out percentage, DP (%) ) (SW ÷ TW) × 100; and fillet yield, FY (%) ) (FW ÷ TW) × 100. Growth or weight decrease during the experimental period was measured with the computation of the following parameters: food conversion ratio, FCR, dry food fed (g) ÷ increase in wet biomass (g); weight gain percent, (final weight - initial weight) ÷ (initial weight) × 100; and specific growth rate, SGR, (lnw2 - lnw1) ÷ (t2 - t1) × 100, where w2 and w1 were the weight in grams at time t2 (end of trial) and t1 (start of trial), respectively. Water Quality Parameters. The principal water quality parameters such as temperature, pH, dissolved oxygen (DO), ammonia-N, nitriteN, nitrate-N, alkalinity, biological oxygen demand (BOD), salinity, total phosphorus, turbidity, total suspended solids (TSS), and total dissolved solids (TDS) were assessed on the water of both systems and analyzed by the Deakin University Water Quality Laboratory (NATA accredited) using standardized methodology routinely implemented in the laboratory. Fatty Acid Analysis. The quantification of fatty acids was conducted as previously reported in our laboratory (9, 13). Briefly, after extraction with chloroform/methanol (2:1v/v) (14) as modified by Ways and Hanahan (15), fatty acids were esterified into methyl esters using the acid catalyzed methylation method (16) and followed by the methods previously used in the laboratory. The internal standard used was 23:0 (Sigma-Aldrich, Inc., St. Louis, MO, USA), and fatty acid methyl esters were isolated and identified using a Shimadzu GC 17A (Shimadzu, Chiyoda-ku, Tokyo, Japan) equipped with an Omegawax 250 capillary column (30 m × 0.25 mm internal diameter, 0.25 µm film thickness, Supelco, Bellefonte, PA, USA), a flame ionization detector (FID), a

-10.5 ( 0.6 b

9039

a Values with the same lower case letter in each row are not significantly different (P > 0.05). An asterisk (*) after the final weight indicates a statistically significant difference (P < 0.05) compared to the initial weight of the fish. b Treatments abbreviations: CWS2, clean water starved for 2 weeks; CWS4, clean water starved for 4 weeks; RWS2, recycled water starved for 2 weeks; RWS4, recycled water starved for 4 weeks; CWF2, clean water fed for 2 weeks; CWF4, clean water fed for 4 weeks; RWF2, recycled water fed for 2 weeks; RWF4, recycled water fed for 4 weeks. c Refer to Abbreviations Used.

1.0 ( 0.1 b 5.5 ( 0.2 d 0.2 ( 0.0 a -13.4 ( 0.9 a 1.2 ( 0.0 c 8.3 ( 0.7 e 0.3 ( 0.0 b -13.4 ( 0.8 a 0.8 ( 0.0 a 4.2 ( 0.5 d 0.3 ( 0.0 b -7.6 ( 0.3 c

Growth

0.80 ( 0.0 a 4.1 ( 0.3 d 0.3 ( 0.0 b FCR c weight gain (%)c SGRc

570.4 ( 7.2 601.7 ( 8.7* 566.7 ( 33.4 c 11.3 ( 0.7d 3.7 ( 0.4 1.6 ( 0.0d 1.8 ( 0.1 c 40.1 ( 0.4 90.4 ( 0.5 ab 561.1 ( 10.5 485.7 ( 9.3* 404.7 ( 20.5 a 3.4 ( 0.2 a 3.0 ( 0.3 1.3 ( 0.0 ab 0.8 ( 0.0 a 38.8 ( 0.7 93.0 ( 0.4 c 557.9 ( 5.0 604.0 ( 1.5* 509.7 ( 32.5 bc 9.9 ( 0.7 cd 3.2 ( 0.4 1.5 ( 0.1 cd 1.8 ( 0.2 c 36.9 ( 2.8 90.4 ( 0.4 ab 557.0 ( 3.0 482.6 ( 6.9* 441.1 ( 30.0 ab 3.5 ( 0.1 a 3.0 ( 0.4 1.3 ( 0.0 a 0.8 ( 0.1 a 38.9 ( 0.6 92.6 ( 0.4 c 570.4 ( 7.2 594.1 ( 7.6 527.9 ( 28.6 bc 10.6 ( 1.1d 3.8 ( 0.5 1.5 ( 0.0 cd 1.8 ( 0.1 c 38.0 ( 0.5 90.1 ( 0.9 a 561.1 ( 10.5 518.2 ( 11.2* 478.4 ( 22.8 abc 4.0 ( 0.3 a 3.6 ( 0.6 1.4 ( 0.0 bc 0.8 ( 0.0 a 38.7 ( 0.5 92.5 ( 0.7 c 557.0 ( 3.0 498.6 ( 3.8* 473.5 ( 31.5 abc 4.9 ( 0.5 a 3.8 ( 0.4 1.4 ( 0.0 bc 1.0 ( 0.1 a 38.5 ( 0.7 92.1 ( 0.4 bc 561.5 ( 17.2 561.5 ( 17.2 475.9 ( 15.9 abc 8.0 ( 0.9 bc 3.2 ( 0.5 1.6 ( 0.0d 1.5 ( 0.1 bc 38.8 ( 26.5 90.2 ( 0.1 a initial weight (g) final weight (g) somatic weight (g) liver weight (g) VFIc Kc HSIc FY (%)c DP (%)c

CWF2 CWS2 initial

557.9 ( 5.0 581.0 ( 4.0* 504.9 ( 30.9 abc 7.4 ( 1.0 b 3.7 ( 0.8 1.5 ( 0.0 cd 1.3 ( 0.2 b 38.2 ( 1.2 91.5 ( 0.6 abc

RWS4 RWS2

RWF2

CWS4

CWF4

4 W purgingb 2 W purgingb biometry

Water Quality Parameters. Nitrite, nitrate, alkalinity, BOD, total phosphorus, turbidity, and TSS were higher in the recycled water (Table 1). The high TDS value in the clean water was due to the salt used for primary prophylaxis, while dissolved solids in the recycled water entirely originated from leached uneaten food and feces. Biometric and Growth Parameters. During the experimental purging, starved Murray cod lost weight. All starved treatments had a significant (P < 0.05) reduction in body weight compared to the start of the trial, while fed fish, apart from RWF2, significantly gained weight (Table 2). The same trend (P < 0.05) was observed for the condition factor (K), the hepatosomatic index (HSI), and dress-out percentage (DP). The fed treatments showed no differences

0 W Purging

RESULTS

Table 2. Biometry and Growth Data (Mean ( SEM) Relative to the Purging Process in Commercial Size Murray Coda

Shimadzu AOC-20i auto injector, and a split injection system (split ratio 50: 1). The temperature program was 150 to 180 at 3 °C min-1, then from 180 to 250 at 2.5 °C min-1, and held at 250 °C for 10 min. The carrier gas was helium at 1.0 mL min-1, at a constant flow. Each of the fatty acids was identified relative to known external standards. The resulting peaks were then corrected by the theoretical relative FID response factors (17) and quantified relative to the internal standard. Flavour Volatile Compounds and Off-Flavor Analysis. A 50 g sample of thoroughly homogenized Murray cod fillet was placed in a 100 mL modified Pyrex bottle fitted with a Shimadzu predrilled rubber septum. The bottle was placed in a water bath and heated to 70 °C. When the temperature was reached, a preconditioned 100 µm polydimethylsiloxane (PDMS) SPME fiber (Supelco) was manually inserted into the Pyrex vial and exposed to the headspace for 60 min. This solidphase microextraction procedure has been chosen after modification of previously reported conditions (18) and several optimizations trials implemented within the laboratory. The fiber was then withdrawn from the sample and immediately desorbed at 270 °C in the injection port of an HP6890 gas chromatograph equipped with a 5973 mass selective detector (Agilent Technologies, Palo Alto, CA). The fiber was left inside the injection port during the entire run to allow removal of all possible residues before the next analysis. The injection port was operated in splitless mode. The head pressure was set to 30 psi of helium for 1 min and then to a constant velocity of 33 cm s-1 for the remainder of the GC run. The initial oven temperature was set at 40 °C for 3 min, then ramped to 200 at 5 °C min-1, and finally ramped to 250 at 50 °C min-1 for a total run time of 40 min. A BPX5 capillary column (30 m × 0.25 mm i.d., 0.25 µm thickness) was used in this study. Each sample was extracted and injected twice, in triplicate, for flavor volatile compounds and for specific geosmin and MIB quantification. The GC was operated in full scan mode for flavor volatile compounds and in selective ion monitoring (SIM) mode for detection and identification of geosmin and 2-methyisoborneol (MIB). Ions at m/z 112, 126, and 182 were monitored for geosmin, and m/z 95, 135, and 168 were monitored for MIB. Identification of compounds was based on mass spectra from library databases (NIST 98, WILEY 275) and known external standards. Data were recorded and analyzed with the Agilent Chemstation Software. The data were calculated as percentage of the total volatile compounds. Because of the different kinetics and partitioning coefficients of different analytes onto the SPME fiber, these analyses were mainly aimed to determine differences between samples rather than comparing different analytes within the same sample. Statistical Analysis. Data are reported as the mean ( SEM (n ) 3). After normality and homogeneity of variance were confirmed, one way analysis of variance (ANOVA) was used to determine differences between means and the multivariate general linear model (GLM) to separate the effects of feeding regime, water, and interaction of the two for data relative to growth, fatty acid, and volatile compounds analysis. Differences were considered statistically significant at P < 0.05. Data were subject to Duncan’s post hoc test where differences were detected for homogeneous subsets. All statistical analyses were performed using SPSS (SPSS Inc. Chicago, Illinois) v.15 for Windows.

J. Agric. Food Chem., Vol. 56, No. 19, 2008

RWF4

Farmed Murray Cod (Maccullochella peelii peelii)

9040

J. Agric. Food Chem., Vol. 56, No. 19, 2008

Palmeri et al.

Table 3. Effects of Feeding Regime, Water Quality, and Time on the Growth, Body Composition, and Recovery Indices of Murray Coda feeding regime weight somatic weight liver VFI fillet HSI K DP FY a

water

feeding regime × water

time

feeding regime × time

water × time

feeding regime × water × time

F value

P

F value

P

F value

P

F value

P

F value

P

F value

P

F value

P

16.206 13.238 150.816 0.628 10.576 137.501 32.564 22.980 0.302

*** ** *** ns ** *** *** *** ns

0.391 0.321 3.531 0.059 0.872 1.1582 2.770 0.137 1.027

ns ns ns ns ns ns ns ns ns

1.504 1.536 4.450 1.582 0.731 2.328 0.932 0.003 0.360

ns ns * ns ns ns ns ns ns

1.993 1.689 8.874 0.330 2.747 5.114 0.104 1.751 0.874

ns ns * ns ns * ns ns ns

3.260 3.057 7.685 0.507 2.710 5.367 7.434 1.288 0.001

ns ns ns ns ns * ** ns ns

0.036 0.007 0.193 0.247 0.068 0.301 0.463 0.808 0.941

ns ns ns ns ns ns ns ns ns

0.579 0.775 1.741 0.037 1.278 4.322 0.009 0.616 1.354

ns ns ns ns ns * ns ns ns

Asterisks denote the level of significance: P *** 0.05) change in fish at time zero and in fish reared in clean water (CWS2, CWF2, CWS4, and CWF4) and in recycled water for two weeks (RWS2). All other treatments recorded a significantly higher lipid content compared to the start of the trial as shown in Table 4, with the highest values recorded for fish fed and starved in recycled water for 4 weeks (RWS4 and RWF4). Fatty acid composition of fillet of fish (expressed as percent of all fatty acids) was significantly modified by the purging strategy implemented (Table 4). Saturated fatty acids (SFA) were significantly higher in RWF4 fish (29.1 ( 0.0%) compared to CWS4 (27.3 ( 0.7%), (the two most extreme treatments), while monounsaturated fatty acids (MUFA) were lowest for CWF4 (33.0 ( 1.6%), compared to all other treatments. Also, CWF4 recorded the highest polyunsaturated fatty acid (PUFA) content (30.7 ( 1.2%), while the lowest value was recorded in RWF4 (28.2 ( 0.2%). RWF4 also recorded the lowest value for the highly unsaturated fatty acids (HUFA; the polyunsaturated fatty acids with three or more double bonds and a chain length of 20 or more carbons).

Farmed Murray Cod (Maccullochella peelii peelii)

J. Agric. Food Chem., Vol. 56, No. 19, 2008

9041

Table 5. Effect of Feeding Regime, Water Quality and Time on the Fatty Acid Composition of Farmed Murray Coda feeding regime F value 14:0 16:0 16:1 n-7 18:0 18:1 n-9 18:1 n-7 18:2 n-6 18:3 n-3 18:4 n-3 20:1 20:4 n-6 20:5 n-3 22:5 n-3 22:6 n-3 SFA MUFA PUFA HUFA n-3 n-6 n-3 HUFA n-6 HUFA n-3/n-6 a

11.661 0.177 0.412 5.578 5.655 4.988 25.183 5.742 0.956 50.846 5.907 12.478 70.871 4.025 0.322 4.966 1.795 19.169 21.466 8.990 108.782 31.158 5.907

** ns ns * * * *** * ns *** * ** *** ns ns * ns *** *** ** *** *** *

water

feeding regime × water

time

feeding regime × time

water × time

feeding regime × water × time

F value

P

F value

P

F value

P

F value

P

F value

P

F value

P

1.278 2.987 1.353 10.177 1.662 2.372 1.735 3.962 2.454 2.397 1.099 16.608 0.108 0.590 2.666 0.964 0.534 2.745 3.232 0.353 7.506 2.373 1.099

ns ns ns * ns ns ns ns ns ns ns ** ns ns ns ns ns ns ns ns * ns ns

13.091 0.572 1.384 1.424 0.193 2.076 0.304 24.397 0.435 5.674 1.757 2.673 2.074 3.444 1.471 3.150 6.712 5.231 13.558 1.091 4.389 4.800 1.757

** ns ns ns ns ns ns *** ns ns ns ns ns ns ns ns * * ** ns ns ns ns

1.278 0.281 1.392 0.029 0.452 0.062 3.045 1.571 0.275 8.170 0.536 2.673 0.001 1.440 0.326 3.345 4.204 1.919 4.662 2.926 0.234 1.806 0.536

ns ns ns ns ns ns ns ns ns * ns ns ns ns ns ns ns ns ns ns ns ns ns

2.161 3.629 0.969 6.790 1.002 0.484 2.045 30.571 1.188 0.227 1.757 1.673 0.338 0.009 4.065 7.684 0.534 0.554 3.121 0.312 6.649 0.169 1.757

ns ns ns * ns ns ns *** ns ns ns ns ns ns ns * ns ns ns ns * ns ns

0.013 0.271 1.253 0.171 0.075 4.779 0.257 0.005 0.022 0.014 0.010 11.506 2.866 1.557 0.199 3.965 3.313 3.803 8.668 0.105 5.091 4.963 0.010

ns ns ns ns ns ns ns ns ns ns ns ** ns ns ns ns ns ns * ns ns ns ns

3.273 0.039 1.432 3.336 0.528 0.142 0.001 32.223 3.202 11.929 0.113 10.220 6.258 0.473 0.133 9.699 5.825 3.272 16.990 0.026 14.960 3.785 0.113

ns ns ns ns ns ns ns *** ns ** ns * * ns ns * * ns ** ns ** ns ns

Asterisks denote the level of significance: ***